partisan bias in economic news: evidence on the agenda ...bias that is correlated with the...
Post on 04-Sep-2020
3 Views
Preview:
TRANSCRIPT
Partisan Bias in Economic News:
Evidence on the Agenda-Setting Behavior
of U.S. Newspapers1
Valentino LarcineseDepartment of Government and STICERD
London School of Economics
Riccardo PuglisiDepartment of Political Science
University of Pavia
James M. Snyder, Jr.Department of Government
Harvard Universityand NBER
November 2010
1We thank Radu Ban, Bob Erikson, Raphael Franck, Chap Lawson, Gabe Lenz,Andrea Prat, Carlo Rosa, Jesse Shapiro, Michiko Ueda and seminar participants atBristol, Cambridge, Columbia, Durham, Rome (Ente Einaudi), LSE, MIT, Milan,Suffolk, the World Meeting of the Public Choice Societies (Amsterdam) for helpfulcomments. We also give a very special thanks to Michael Naber for making thedata collection possible.
Abstract
We study the agenda-setting political behavior of a large sample of U.S. newspapers duringthe 1996-2005 period. Our purpose is to examine the intensity of coverage of economicissues as a function of the underlying economic conditions and the political affiliation of theincumbent president, focusing on unemployment, inflation, the federal budget and the tradedeficit. We investigate whether there is any significant correlation between the endorsementpolicy of newspapers, and the differential coverage of bad/good economic news as a functionof the president’s political affiliation. We find evidence that newspapers with pro-Democraticendorsement pattern systematically give more coverage to high unemployment when theincumbent president is a Republican than when the president is Democratic, compared tonewspapers with pro-Republican endorsement pattern. This result does not appear to bedriven by the partisanship of readers. We find similar but less robust results for the tradedeficit. We also find some evidence that newspapers cater to the partisan tastes of readersin the coverage of the budget deficit. We find no evidence of a partisan bias – or at least of abias that is correlated with the endorsement or reader partisanship – for stories on inflation.
1. Introduction
News provided by the mass media are the most important source of information on
public affairs in modern democratic societies. Hence, media outlets play a fundamental role
in keeping the public informed on the decisions of their political representatives, as well as
on issues and events that are relevant to public decision-making. Time and space available
being limited, journalists exercise a considerable degree of discretion on the topics covered
and the tone of the reports. It would therefore not be surprising if the political views of
individual editors and journalists were reflected in news reported in the mass media.
One of the most important claims about news in the mass media is the agenda-setting
hypothesis. The idea is that editors and journalists have a large degree of freedom in deciding
what is newsworthy and what is not, and these choices influence the perception of citizens
about which issues are relevant and to what extent. Cohen [1963] stated it eloquently:
the press “may not be successful much of the time in telling people what to think, but it is
stunningly successful in telling its readers what to think about.”1 The exploitation of agenda-
setting power is potentially one of the most harmful behaviors by news media, especially if
they use this power to suppress information. The reason is that it is difficult for consumers
to distinguish the scenario “I did not see any news about X today because nothing important
happened regarding X” from the scenario “I did not see any news about X today because,
although something important happened, the media decided not to publish it”. Theoretical
models by Anderson and McLaren [2009], Baron [2006], Bernhardt et al. [2008], Besley and
Prat [2006] and Puglisi [2004] incorporate precisely this source of media bias, and show how
this can lead to suboptimal public policy decisions.
In this paper we try to gauge the extent of agenda bias on economic issues for a large
number of U.S. newspapers over the period 1996-2005. The logic of our approach is as
follows. Consider the issue of unemployment and suppose that the incumbent president is
a Democrat. Suppose also that some newspapers have a partisan bias and wish to increase
or decrease the popularity of the president. When unemployment is high or rising (i.e.,
1Beginning with the seminal contribution of McCombs and Shaw [1972], there is a vast literature incommunication studies on agenda-setting effects. See also Iyengar et al. [1982], and Iyengar and Simon[2000].
1
when the underlying circumstances are bad) Republican leaning newspapers should devote
more coverage to that issue than Democratic newspapers. The opposite should occur when
unemployment is low or falling (i.e., when circumstances are good).2
We apply this logic to four key economic issues: unemployment, the budget deficit,
inflation, and the trade deficit. These issues are not only important, but we can also match
the coverage with actual economic figures. We collected data on the number of news stories
on each of these issues appearing in a large sample of newspapers, using the NewsLibrary
and Factiva electronic archives.
For each issue we construct a measure of newspaper bias based on the differential sen-
sitivity of coverage to the underlying economic figures as a function of the party of the
president. We investigate whether this measure is systematically related to supply and/or
demand factors. As a proxy for the explicit partisan leaning of owners and editors of a given
newspaper, we use the relative propensity to endorse Republican or Democratic candidates
across a large sample of elections. As a proxy for the partisan leaning of a newspaper’s
readers, we use circulation-weighted voting data at the county level.
For unemployment, we find robust evidence of a correlation between intensity of news
coverage and the partisanship of endorsements. More precisely, we find that newspapers
with a pro-Democratic endorsement pattern systematically publish fewer stories about un-
employment when the national unemployment rate is high and the president is a Democrat
than when the national unemployment is equally high and the president is a Republican.
The size of the estimated effects is nonnegligible, especially when expressed in relative terms.
When the unemployment rate was one percentage point above the average, newspapers with
a strong propensity to endorse Republican candidates reacted with about 10% fewer articles
per month under Bush than under Clinton. For the same one percent increase, newspapers
with a strong pro-Democratic endorsement policy published around 7% more stories on un-
employment under Bush than under Clinton. Note that we do not make any claims about
the absolute biases of newspapers, but only their relative positions.
2We focus on the political affiliation of the incumbent president, because national economic conditionsare most closely associated with the popularity and vote of the president, while there is no robust evidencesuggesting that the economy has any significant effects on congressional elections (Fair 1978, Tufte 1978,Hibbs 1987, Erikson 1989, 1990, MacKuen et al. 1992).
2
With respect to readership, we find mixed results. While newspapers more heavily sold in
Democratic areas tend to give more coverage to high unemployment under Bush than under
Clinton as compared to those sold in Republican areas, this correlation is not significant,
irrespective of whether we control for endorsement partisanship. On the other hand editorial
partisanship is still significant when both variables are included, and it is robust to a large
set of controls. These results suggest that agenda bias on unemployment is more related to
the partisan position of owners and editors than to the partisanship of readers.
The situation is reversed for coverage of the budget deficit. In this case newspapers ap-
pear to cater to the partisan tastes of their readers, but we find no statistically significant
correlations with endorsement partisanship. Newspapers that are more heavily sold in Re-
publican areas systematically devote more coverage to the budget deficit when the deficit is
high and the incumbent President is a Democrat than when he is a Republican, as compared
to newspapers sold in Democratic areas. The size of the estimated effect is again nonnegligi-
ble, and larger than that found for unemployment, at least relative to the average amount of
coverage devoted to the issue. When the budget deficit was one percentage point above the
average, newspapers circulating among strongly Republican readers published about 20%
more articles per month under Bush than under Clinton. For the same one percent increase,
newspapers with a strongly Democratic readership reacted with about 28% more stories on
the budget deficit under Bush than under Clinton. This result holds even when controlling
for endorsement partisanship but is less robust than our result on unemployment, since it
loses significance when controlling for the lagged instead of the contemporaneous level of the
budget deficit.
We find results for the coverage of the trade deficit that are qualitatively similar to those
on unemployment, but less robust. Finally, we find no systematic relationships between
inflation coverage and either editorial partisanship or reader partisanship.
To sum up, we find that both supply-side and demand-side forces matter, although our
results on the role played by the supply side are somewhat more robust.
Importantly, the issue for which we find the most robust correlations – unemployment – is
also the most salient of the four during the time period studied. This is clear from the relative
amount of coverage devoted to the issues by the newspapers themselves, and also from survey
3
data. In our sample the breakdown of coverage is as follows: 50% of the newspaper articles
discuss unemployment, 37% discuss inflation, 9% discuss the budget deficit, and only 3%
discuss the trade deficit.3 In the American National Election Studies from the period 1992-
2004, nearly 10% of respondents cited unemployment as the “most important problem facing
the nation.”4 By comparison, fewer than 0.5% of respondents mentioned inflation and, even
counting generously, only about 1.5% of respondents mentioned trade issues.5 Unfortunately,
the survey data does not allow us to separate the government deficit from the general issue
of government spending. It is likely that survey respondents did not perceive inflation to be
a significant problem because during the period under study the inflation rate was generally
low. During the 1992-2005 period, the highest inflation rate (CPI) was 3.4%.6
Our paper makes three contributions to the economics literature on media bias. First,
we focus on the intensity of coverage across issues, rather than tone. Theoretically, it is
arguable that intensity of coverage – especially, suppression of coverage – is more important
than tone, because it poses a particularly difficult inference problem for citizens. Second,
we focus on important economic topics that are relevant to all citizens and policy makers.
Third, although we do not do this here, it is straightforward to apply our measurement
strategy to different countries and time periods.
Finally, a salient feature of our approach is that we code newspaper articles through an
automatic keyword search, instead of a human-based content analysis. One advantage of
this procedure is that, by definition, it is not intensive in the usage of human capital. Its
low cost means that it can be used to gather data on a large number of news outlets for
a long time span, restricted only by availabilities in digital archives. More importantly, an
automatic search is easily replicated, as it is based on known set of words and/or sentences
that are used as classifiers.7
3Of course, some articles discuss more than one issue.4Overall, crime was mentioned most often, and unemployment second most often.5Only 0.33% of respondents mentioned the trade deficit specifically, and more respondents mentioned
“international competitiveness” or “outsourcing”, which might be treated more appropriately as employmentissues.
6With respect to inflation, there is another reason to suspect that partisan bias is less salient. Theindependence of the Federal Reserve makes it difficult for the public to establish links between presidentialpolicies and inflation.
7As pointed out by Antweiler and Frank [2005], automated procedures of text classification have thefurther advantage of reducing the “degrees of freedom” available to the researcher in the choice of the media
4
The paper is organized as follows: in the next section we briefly review the related
literature, while in section 3 – as a case study on the relevance of supply side factors in
determining slanted coverage of economic news– we discuss the succession of Otis Chandler
as publisher of the family-owned Los Angeles Times in 1960. In section 4 we describe the
data and the empirical strategy, in section 5 we present the main results, and in section 6
we present various robustness checks. Section 7 concludes.
2. Related Literature
In the theoretical literature there are three approaches to modelling media bias. In the
first approach, citizens have preferences directly over the ideological content of the news they
consume, and media outlets cater to these preferences (Mullainathan and Shleifer 2005). In
the second approach media bias takes the form of “pandering” to citizens’ prior beliefs, in
order to maintain a reputation for reliable reporting (Gentzkow and Shapiro 2006). In the
third approach, citizens seek information needed to evaluate policies or politicians. This
information is assumed to come from media outlets, and these outlets may suppress or skew
the information (Anderson and McLaren 2009, Baron 2006, Bernhardt et al. 2008, Besley
and Prat 2006, and Puglisi 2004). As noted in the introduction, in this case it may be difficult
even for highly rational citizens to completely undo the malicious effects of news bias. In
the first two approaches the bias is driven by demand-side forces. In the third approach bias
is driven by supply-side factors. One such factor is the desire of politicians to suppress news
that will hurt them. Another is ideological consumption by owners, editors, and journalists.
Demsetz and Lehn [1985] discuss the “amenity potential” for owners of media firms, and
find evidence that the scope for such consumption is large.8
The empirical studies on measuring bias can also be divided into three groups. One type
outlets to be included in the sample. Gentzkow and Shapiro [2010] and Puglisi and Snyder [2008] also adopta keyword-based approach.
8Other studies such as Gilens and Hertzman [2000], Puglisi and Snyder [2008] and Durante and Knight[2010] find evidence that media content is significantly correlated with supply-side factors. Also, the in-centives to bias media content should be stronger if the bias has a persuasive effect on readers. A numberof studies show that this is the case (e.g., DellaVigna and Kaplan 2007, Gerber et al. 2009, and Knightand Chiang 2008). In addition, Corneo [2006] and Petrova [2008a] provide formal models on the effects ofwealth concentration on media behavior and policy choices. Ellman and Germano [2009] and Petrova [2008b]focus on the role played by advertisers and interest groups. See also the recent surveys by Della Vigna andGentzkow [2010] and Prat and Stromberg [2010].
5
focuses on the explicit political behavior of newspapers, analyzing endorsements of candidates
or ballot propositions (e.g., Ansolabehere et al. 2006, Puglisi and Snyder 2009). A second
type measures the implicit political behavior of media outlets, analyzing the language they
use or the sources they cite in their news stories (e.g., Gasper 2007, Gentzkow and Shapiro
2010, Groseclose and Milyo 2005). The idea is to compare the words, phrases or sources
used by the media with those used by politicians. Outlets that employ language or sources
that are used mainly by Republican (Democratic) politicians are then classified as relatively
conservative (liberal). The third type also measures the implicit political behavior of the
media, but focuses on the amount of coverage devoted to various issues, that is, on agenda-
setting (e.g., Puglisi 2006, Puglisi and Snyder 2008). The idea is to analyze how the behavior
of newspapers varies as the partisan identity of the sitting president (or main national leader)
varies. For example, Puglisi and Snyder [2008] study political scandals. A newspaper is
classified as relatively conservative (liberal) if it devotes relatively more attention to scandals
involving Democrats (Republicans).9
Our paper provides a new measure of the third type of bias.
3. The case of the Los Angeles Times
We begin with a case study, which illustrates our approach and provides initial evidence
that supply-side factors may account for some newspaper behavior. The case involves the
succession of Otis Chandler in 1960 as publisher of the Los Angeles Times. The Chandler
family owned the LA Times from 1884 to 2000. Prior to 1960 it was widely perceived as
having a conservative, pro-Republican bias. Chandler sought to change this, and transform
the paper into a credible rival of the New York Times.10
Figure 1 shows the time-series variation in the propensity of the LA Times to endorse
9Lott and Hassett [2004] shares features of the second and third groups. They analyze newspaper coveragewhen official data about various economic indicators are released. They code the “tone” – positive or negative– of newspaper headlines, and relate this to the partisanship of the sitting president. Larcinese [2007] studiesanother type of bias – the propensity for newspapers in the UK to overprovide news that is of interest toaudiences that are more valuable to advertisers. Stromberg [2004] provides a formal model that rationalizesthis type of behavior.
10He succeeded on some dimensions. For example, the daily circulation of the LA Times went fromapproximately 500,000 in 1960 to over 1,000,000 in 1976. The newspaper also won four Pulitzer Prizes inthe 1960s, which was more than it had won in the previous 90 years combined. See Halberstam [2000] for ahistorical account of the changes.
6
Democratic candidates in California statewide and congressional elections, together with the
average yearly share of the Democratic vote in presidential, senatorial and gubernatorial
elections in California. In the 1960s, after Otis Chandler took control, there was a steep
increase in the propensity to endorse Democratic candidates.11 This was not matched by
a comparatively rapid surge in the Democratic vote. This suggests that there was a large
change in the “tastes” of the LA Times editors, with Otis Chandler being much less pro-
Republican than his predecessors.
Figure 2 presents the salient patterns regarding news coverage. The top two scatter
plots in the figure show the relationship between the actual unemployment rate and the
relative frequency of unemployment stories in the LA Times, before and after 1965.12 In each
graph, coverage-unemployment combinations under a Democratic (Republican) president are
indexed by a one (zero). The bottom two graphs parallel the top graphs, showing the same
relationship for the inflation rate. The figures also show the estimated regressions lines
relating the economic variables and coverage, as a function of the political affiliation of the
incumbent President. The two scatter plots on the left show that before 1965 the LA Times
systematically gave more coverage to high unemployment and inflation – i.e., more coverage
to bad economic news – under Democratic presidents than under Republican presidents.13
This is evidence of a pro-Republican bias. On the other hand, the two graphs on the right
show that after 1965 there is no systematic difference in the slopes under presidents of
different parties. That is, after Otis Chandler took over as publisher, the pro-Republican
bias exhibited by the LA Times disappeared.
Ideally, we would like to expand this single case study into a large-scale analysis. However,
this would require data on a large sample of newspapers over a long period of time, in which
there was a significant number of changes in ownership or management. Currently, collecting
the necessary data would be an extremely time-consuming and expensive task. Such a study
might become feasible in the near future, as historical archives for more newspapers become
11Other members of Chandler family kept some influence and, in spite of a clearly more liberal leaning,the LA Times remained for some time a Republican newspaper. It endorsed Nixon in 1960 and 1968 andmildly endorsed Goldwater in 1964.
12The data on the number of stories is from Proquest. We discuss how the data was collected in moredetail below.
13This is confirmed by proper difference-in-differences regressions, available upon request from the authors.
7
available on-line. For the present, we hope to satisfy the reader with an analysis of a large
number of newspapers over a short period of time. This study necessarily focuses on cross-
sectional variation rather than variation over time.
4. The panel data and empirical strategy
We collected data from the NewsLibrary electronic archive, recording the monthly number
of hits on unemployment and inflation, and the quarterly number of hits on the federal
budget deficit and the trade deficit.14 First, we conducted a number of preliminary searches
to identify a set of keyword search strings which produced a relatively low number of false
positive and false negative hits. Then we ran automated searches, retrieving the number of
articles containing the selected keywords for each topic in each newspaper and time interval.
Overall, we collected data on 140 U.S. newspapers for which electronic archives dating back
to 1996 are available to be searched through NewsLibrary. We use the newspapers’ own
archives to add data on the Los Angeles Times and the Chicago Tribune, and the Factiva
archive for the New York Times.15
In this section we will first present some summary statistics of the economic news data,
and describe how we use it to compute a measure of partisan coverage. We then illustrate the
procedure used to recover the endorsement propensity of the various newspapers. We do the
same for the measure of reader partisanship. Finally we investigate the simple correlation
between our measure of agenda bias and either endorsement or reader partisanship. This
illustrates our empirical strategy, in a less rigorous but more intuitive fashion. We will then
be ready to present our panel specification.
14The official macroeconomic figure is made available to the public monthly for the unemployment andthe inflation rate, and quarterly for the two deficits.
15We conducted some ex-post checks and detailed reading of random samples of articles, focussing onunemployment. We used various sampling strategies, all providing broadly similar results. In one instancewe (1) randomly chose 10 newspapers, then (2) randomly chose 10 months, then (3) randomly chose 3 dateswithin the months. We obtained 229 hits of which 203 (88.6%) were ”good” hits and 26 false positives. Ofthe good hits, 147 were about unemployment level or unemployment rate (local, state, or national but notforeign unless there was also a comparison with the U.S.): this is 72.4% of the good hits (147 out of 203). Ofthe remaining good hits, 6.4% were about layoffs, 5.4% about personal stories, 3.4% about unemploymentinsurance policy, 2.5% about the difficulties of being unemployed generally. Of the 26 false positives, 12involved unemployment in foreign countries (46.2%). Other sampling methods provided broadly similarpatterns.
8
4.1. Economic news data
The key variables in our analysis are the values of the four underlying economic indicators,
and the amount of newspaper coverage devoted to the four economic issues. Since newspapers
vary greatly in size cross-sectionally (total number of pages, stories, and words), and can
also vary in size over time, our dependent variable is the relative frequency of stories in each
newspaper during each time period about a given economic issue.
More formally, we focus on nijt = hi
jt/Hjt – i.e. the relative frequency of articles in
newspaper j at time t regarding issue i, where i ∈ {unemployment, inflation, budget deficit,
trade deficit}. The numerator hijt is the count of stories appearing in outlet j at time t which
contain the keywords related to issue i. Table 1 reports the keywords that we use.16 The
denominator Hjt is the number of stories in newspaper j and time period t in which the
word “and” appears, which we use as a proxy for the total number of stories.
To get an initial sense of the variation in newspaper coverage, consider the following. Let
EV it be the value of the economic figure regarding issue i at time t.17 For each newspaper j
and each economic issue i, we run a separate OLS regression:
nijt = αi
j + βi1j EV
it + βi
2j ∆EV it + γi
j DPt + δij (EV i
t ·DPt) + λij ln sjt + εijt (1)
where DPt is a dummy variable indicating that the incumbent president is a Democrat. In
addition, we control for the change in the economic variable of interest (month by month for
unemployment and inflation, quarter by quarter for the budget and the trade deficit). We
also control for the logarithm of the total number of articles in each newspaper at time t,
sjt. The coefficient δij represents the difference in how newspaper j reacts to bad economic
news when the president is Democratic compared to when the president is a Republican.
Positive values indicate that the newspaper is more reactive to bad economic news when the
incumbent president is a Democrat.18
16A potential concern is that all the variation in the coverage of economic news might be driven byeditorials. Thus, we also ran the searches excluding the words “editorial” or “editor”. We explore therobustness of our results to this narrower definition of coverage below.
17Table 2 displays summary statistics of the relative frequency of stories and the economic figures ofinterest for the 1996-2005 period.
18If we had data for a period long enough to cover numerous presidents, it would be possible to treatthis interaction term as a measure of the absolute pro-Republican bias of a newspaper. However, given theshort time span available, the time series variation by itself could easily be misleading. In particular, othernewsworthy events and issues could be crowding out economic news more in some years than others.
9
There is considerable variation in the differential coverage of the four economic issues we
study. Consider unemployment. The Fresno Bee lies at one extreme, with δj = −0.87. That
is, given a one-percentage-point increase in the unemployment rate, the Fresno Bee would
devote almost one percent fewer of their stories to unemployment under Clinton than under
Bush. In relative terms this is a fairly large difference, since the newspaper only devotes 1.35
percent of its stories to unemployment on average. The Bismark Tribune lies at the opposite
extreme, with δj = 0.46. A one-percentage-point increase in the unemployment would lead
this paper to print one-half of one percent more stories under Clinton than under Bush (on
average). Most newspapers are noticeably more centrist, including almost all of the largest
newspapers. For example, the estimated δj is −0.075 for the New York Times, −0.19 for the
Los Angeles Times, and 0.136 for the Detroit Free Press.
4.2. Endorsement and readership data
We collected endorsement data for 102 newspapers. For 85 newspapers the data is from
Ansolabehere et al. [2006]. We supplement this with data on 17 additional newspapers
searched via the NewsLibrary archive. For the remaining 38 newspapers in our sample, in
some cases the newspaper has an explicit policy not to endorse candidates for political offices
(e.g. the Deseret News in Salt Lake City, the Orange County Register, and the Colorado
Springs Gazette). In addition, many smaller ones do not bother to make endorsements, even
though they may not take an explicit editorial stance on the subject. Table A1 in the online
appendix lists the newspapers with endorsement data, together with the chain to which they
belong, if any.
Following Ansolabehere et al. we can calculate the propensity of each newspaper to
endorse one of the parties during electoral campaigns. We used a linear regression model
to estimate the “partisan bias” in endorsement behavior. Let k index offices, let j index
newspapers and let t index years. Let
Ekjt =
1 if newspaper j endorses Democrat for office k in year t
−1 if newspaper j endorses Republican for office k in year t
0 if newspaper j makes no endorsement for office k in year t
measure the endorsement behavior by each newspaper that makes an endorsement (or an
10
explicit refusal to endorse) in a race.19 Also, let
Iijt =
1 if Democrat for office k in year t is only incumbent
−1 if Republican for office k in year t is only incumbent
0 if otherwise
measure the incumbency status of the candidates in each race.20 Finally, we use previous
electoral experience to measure non-incumbent quality. Specifically, define a “high-quality”
candidate as a candidate who currently holds a U.S. House seat or an elected statewide office
other than the office sought. Let
Qijt =
1 if Democrat for office i in year t is only high quality non-incumbent
−1 if Republican for office i in year t is only high quality non-incumbent
0 otherwise
We estimated the following linear model for the period 1992-2006, exploiting the panel
nature of the data21
Eijt = NEj + θt + β1Iijt + β2Qijt + εijt (2)
The newspaper-specific fixed effects, NEj, capture newspapers’ partisanship.22 Positive val-
ues indicate a propensity to endorse Democratic candidates and negative values a propensity
to endorse Republican candidates.
A few newspapers exhibit strong partisan biases in their endorsement behavior. For
example, the estimated NEj for the Florida Union is −0.75, and the estimated NEj for the
Sacramento Bee is 0.62. The Washington Times is something of an outlier, with an NEj
of −1.14. Overall, however, most newspapers appear to be relatively centrist. More than
half of the newspapers have estimate NEj’s between −0.35 and 0.24. The NEj’s of a few
prominent newspapers are as follows: New York Times = 0.50, Los Angeles Times = 0.27,
Washington Post = 0.21, and Chicago Tribune = −0.24.
As a proxy for the average political position of readers of a given newspaper j, we weight
the average Democratic vote in presidential, senatorial and gubernatorial elections in each
19There are a few cases in our sample where a newspaper endorsed both candidates in a race. We dropthese from our analysis.
20After redistricting there are some U.S. House races with two incumbents running, in which case Iijt = 0.There are a few such cases in our sample. If we drop them the results are unchanged.
21The panel is unbalanced, since we do not have endorsement data on some newspapers in the earlieryears.
22The model also includes year fixed-effects, θt, to capture partisan tides.
11
county during the time period by the relative sales of that newspaper in that county. We
call this variable “reader partisanship,” and denote it by NRj. It is important to note that
since we do not have individual level data on readership, NRj is not equal to the actual
partisanship of newspaper j’s readers. Instead, it measures the partisanship of the voters in
newspaper j’s market area.
Reader partisanship appears to be less concentrated than endorsement partisanship, and
larger newspapers are sold in Democratic and Republican areas as well, rather than being
concentrated in moderate areas. Not surprisingly, the NRj’s for the six largest newspapers
in our sample are larger than one half, suggesting that their readers tend to vote Democratic
more than half of the time.
Ex ante, one might be concerned that endorsement and reader partisanship are so corre-
lated that it is very difficult to tell one from the other. Figure 3 displays a scatter plot of the
endorsement partisanship NEj against readers’ ideology NRj for our sample of newspapers,
together with the estimated regression line. As expected, there is a statistically significant
correlation between the partisan stance on the demand and on the supply side in the cross
section. But the correlation is only 0.23, which is hardly overwhelming. The Washington
Times is a noticeable outlier, but even dropping this newspaper the correlation is only 0.30.
Evidently, there is a substantial amount of “slack” between the partisan positions of news
consumers and news providers.
4.3. Specification
Here, we describe our main specification, which exploits the panel nature of our data.
Consider the relationship between news coverage and endorsement partisanship first. For
each of the economic outcomes i ∈ {unemployment, inflation, budget deficit, trade deficit},
we estimate the following model:
nijt = αi
j + ζ it + βi
j EVit + γi
j DPt + φi(EV it ·DPt · NEj) + λi ln sjt + εijt (3)
where EV it is the underlying economic outcome variable; DPt is a dummy variable indicating
that the president at time t is a Democrat; NEj is the estimated newspaper-specific endorse-
ment propensity from equation 2 above; sjt is the logarithm of the total number of articles
in newspaper j at time t; αij is a newspaper-specific fixed effect on economic issue i; and ζ i
t
12
is a time-specific fixed effect on issue i. This specification is quite general, as it allows each
newspaper to react differently to the party of the president (captured by γij DPt), and also
to the underlying economic variables (captured by βij EV
it ). That is, it allows newspapers
to react differently to DPt and EV it not only as a function of their endorsement partisan-
ship but also as a function of any other (fixed) unobserved newspaper characteristics. The
main coefficient of interest in terms of relative bias is φi, the coefficient on the three-way
interaction term between the economic variable EV it , the party of the president DPt, and
newspaper partisanship NEj. A negative value of φi implies that newspapers that tend
to endorse Democratic candidates have a relatively pro-Democratic agenda-setting bias on
economic issue i, compared to newspapers that tend to endorse Republican candidates.
We use the same type of specification to investigate the relationship between the bias in
coverage and reader partisanship, by replacing NEj with NRj in equation (3). In addition,
we explore the role of demand-side and the supply-side partisanship simultaneously, by
including the three-way interactions for both NEj and NRj in the same specification.
To account for the possibility that the fixed effects may not absorb the entire within-
newspaper correlation in the error term, we run all regressions clustering the standard errors
by newspaper. We also conduct a variety of robustness checks on the baseline specification,
which are presented and discussed in section 6.
One especially important check is to compute bootstrapped standard errors, since the
model includes generated regressors (Pagan 1984, Murphy and Topel 2002). This involves
re-sampling from the endorsement data to generate new NEj’s in each iteration of the boot-
strap. Also, since we are concerned about within-newspaper correlation in the error term,
we use cluster-sampling in implementing the bootstrap.23 It turns out that the bootstrapped
standard errors are actually slightly smaller than the ordinary standard errors, at least for
the main coefficients of interest (φi). Thus, we present the more conservative standard errors
in our baseline table, and report the results of the bootstrap in the robustness section (in
Table 5).
Other robustness checks include: adding control variables, as well as three-way interac-
tion terms of these controls with EV it · DPt; using lagged values of the economic variables
23See Cameron et al. [2008] for an extensive discussion of cluster-bootstrap techniques.
13
rather than contemporaneous values; including local economic conditions alongside national
conditions; using changes in economic conditions rather than levels. Finally, we also con-
sidered several alternative specifications in addition to equation (3). We do not present
the results for these specifications, but they are reported in a working paper version of this
article.24
After matching with all the explanatory variables, our final sample consists of 101 news-
papers for the period 1996-2005. We exclude the Washington Times from all of our regression
analyses, because it is an extreme and influential outlier, and we do not want our results to
be unduly affected by one newspaper.
5. Results
Table 3 presents the results. There are four panels, one for each of the economic variables.
For each economic variable there are three columns. The first column focuses on newspaper
endorsement partisanship, the second focuses on reader partisanship, and the third jointly
considers newspaper endorsement and reader partisanship. The first two rows of each panel
present the estimates of the φi’s, the coefficients on the main variables of interest, i.e., the
three-way interaction terms in equation (3). The third row reports the estimates of λi, the
coefficient on the log of the total number of articles. Clustered standard errors are reported
in brackets below each coefficient.
First, consider the endorsement partisanship columns. The results in the first panel show
that the three-way interaction between the level of the unemployment rate, the Democratic
President dummy and the Democratic endorsement variable has the expected negative sign
and is significant at the 5%. This is evidence of a bias in coverage that is significantly
correlated with the editorial stance of newspapers as measured by endorsements. Newspapers
with a pro-Democratic-endorsement pattern, compared to pro-Republican newspapers, give
significantly less coverage to unemployment in times of high unemployment under Clinton
than under George W. Bush. For the other three economic variables the three-way interaction
terms are all small, although for the trade deficit the coefficient is significant at the 10%.
Next, consider the reader partisanship columns. Here, the estimated three-way interac-
24See Larcinese et al. [2007].
14
tion term is relatively large and significant at the .05 level for the budget deficit issue. This
is evidence of a bias in coverage that is significantly correlated with the partisan leaning
of readers. Newspapers more heavily sold in Democratic counties are significantly more re-
active to a larger budget deficit under George W. Bush than under Clinton, compared to
newspapers sold in Republican counties. For the other three economic variables the point
estimates of the three-way interaction are all negative, but they are never significant even
at the 10% level.
Finally, consider the third column in each panel, where endorsement and reader partisan-
ship are both included in the model. The results basically confirm those in the columns 1 and
2. The differential coverage of unemployment is significantly correlated with endorsement
partisanship, but not with reader partisanship. On the other hand, the differential coverage
of the budget deficit is significantly correlated with the average partisan leaning of readers
but not with that displayed by editors through their endorsements. For the trade deficit
there is again some weak evidence of a significant three-way interaction with endorsement
partisanship. In fact, the three-way interaction is no longer significant when dropping the
New York Times from the sample. The coverage of inflation is not significantly correlated
with either endorsement or reader partisanship.
How large are the effects shown in Table 3? Table 4 presents some simple comparisons,
for the most significant coefficients. In the top panel (panel [A]) we focus on the coverage
of unemployment, stratifying newspapers by their endorsing behavior. In panel [B] we focus
on coverage of the budget deficit, stratifying newspapers by their readers’ partisanship. In
panel [C] we focus on the trade deficit, stratifying newspapers again on the basis of their
endorsing behavior.
More precisely, in panel [A] we group newspapers into quintiles on the basis of their
endorsement patterns. We refer to newspapers belonging to the first, third and fifth quin-
tile in the endorsement distribution, as, respectively, Pro-Republican Endorsing, Neutral
Endorsing, and Pro-Democratic Endorsing. For each group of newspapers we compute the
difference between the average change in the predicted number of stories under Clinton and
the average change in the predicted number of stories under George W. Bush, given a 1
15
percentage point increase in unemployment.25 Thus, for example, given a 1% increase in the
unemployment rate, a Pro-Republican Endorsing newspaper will print 0.056 more stories on
unemployment if the president is a Democrat than if the president is a Republican. This
represents a difference of about 10 percent, since these papers only print an average of 0.576
stories on unemployment per month. By contrast, given a 1% increase in the unemployment
rate, a Pro-Democratic Endorsing newspaper will print 0.051 fewer stories on unemployment
if the president is a Democrat than if the president is a Republican, a difference of about 7
percent.
In panel [B] we group newspapers into quintiles on the basis of their reader partisan-
ship scores. We refer to newspapers belonging to the first, third and fifth quintile in the
endorsement distribution, as, respectively, Pro-Republican Readers, Neutral Readers, and
Pro-Democratic Readers. For each group of newspapers we compute the difference between
the average change in the predicted number of stories under Clinton and the average change
in the predicted number of stories under George W. Bush, given a 1% increase in the bud-
get deficit.26 Again, the magnitudes are substantively meaningful. Given a 1% change in
the deficit, newspapers with Pro-Republican Readers would react by publishing 0.023 more
stories under Bush than under Clinton (a difference of about 20 percent with respect to
the average number of stories), while newspapers with Pro-Democratic Readers would more
strongly react by publishing 0.035 more stories under Bush than under Clinton, a difference
of about 28 percent.
Finally, panel [C] shows that – given a 1% increase in the trade deficit – a Pro-Republican
Endorsing newspaper will print 0.007 more stories on the issue if the president is a Democrat
than if the president is a Republican, a difference of about 20 percent.27 On the other
hand, given a 1% increase in the deficit, a Pro-Democratic Endorsing newspaper will print
0.006 fewer stories on trade deficit if the president is a Democrat than if the president is a
Republican, a difference of about 10 percent.
25To put the 1 percentage point change in perspective, the standard deviation of the unemployment rateduring the period under study was 0.672.
26The standard deviation of the budget deficit during the period under study was 1.936.27The standard deviation of the trade deficit during the period is 1.579.
16
6. Robustness checks
In this section we present the results of a variety of robustness checks. The estimates are
all shown in Table 5.28 As in Table 3 above, we focus on the main parameters of interest,
the coefficients on the relevant three-way interaction variable, φi.
The bottom line from Table 5 is that the estimates in Table 3 are quite robust. Consider
first the unemployment issue. In all cases, the estimates of φi for newspaper endorsement
partisanship are similar to those in Table 3 (in the range -0.11 to -0.13) and statistically
significant at the .05 level. Similarly, for the budget deficit issue the estimates of φi for
reader partisanship are in all but one case similar to those in the baseline specification (in
the range of -0.05 to -0.06) and statistically significant at the .10 or .05 level. The results
for the trade deficit and inflation are also generally similar to those in Table 3.
In panel [A], in order to deal with endorsement partisanship as a generated regressor, we
estimate the standard errors using a bootstrap. Since we must cluster our standard errors by
newspaper, we employ cluster-sampling in the bootstrap, and follow the bootstrap-se pro-
cedure, as described in Cameron et al. [2008]. More precisely, in the first stage we estimate
endorsement scores by resampling newspapers as clusters (not the single endorsements). In
the second stage we then estimate our baseline specification 3 with clustered standard errors.
We run 500 replicas for each regression and extract bootstrapped standard errors from the
resulting distribution. As mentioned above, those standard errors are slightly smaller than
the ones found with our baseline specification.
It is not clear a priori whether newsworthy economic events are more correlated with
contemporaneous values of the relevant economic figures, or lagged values. Government
agencies such as the Bureau of Economic Analysis and Bureau of Labor Statistics can only
publish lagged values of macroeconomic variables. However, newspapers do not only report
on the release of official data – which are related to what happened in the past – but also
on contemporaneous events, which may be correlated with the current value of the relevant
macroeconomic figure. For example, with respect to unemployment, there are sometimes
news stories about large layoffs in a given sector or by a particular firms, or reports of large
28Full results for these robustness checks are available in the online appendix to the paper, more preciselyin Tables A2, A3 and A4.
17
current spikes in applications at local unemployment agencies.
Panel [B] shows the results when we use lagged values of the economic variables of
interest instead of contemporaneous values. More precisely, we use the previous month’s
unemployment rate and inflation rate, and the previous quarter’s budget deficit and trade
deficit. For unemployment, the results are similar to those in Table 3. For the budget deficit,
however, they are not. The three-way interaction with reader partisanship is only about half
as large as that when we use contemporaneous values, and no longer significant even at the
.10 level.
Newspapers typically have a locally concentrated readership that cares about local events,
and local aspects of common phenomena. Since there is noticeable variation in unemploy-
ment across regions and states, the local unemployment rate in an area or state may represent
a newsworthy issue. This can potentially introduce an omitted variables bias. The concern
is that, in Democratic-voting areas, the local unemployment rate could be systematically
lower than its average when the incumbent president is a Democrat, because of public job-
creating projects being targeted to the area. Since the political partisanship of potential
readers in the area where a newspaper sells is positively correlated with its endorsement
policy (see Figure 3), it is possible that the less intense coverage of high unemployment by
Democratic-leaning newspapers under a Democratic president is driven by the fact that the
local unemployment rate is lower in those areas where the newspapers are sold. This would
not indicate a partisan bias trickling down from the editorial page to the economic news
section, but simple reporting on local economic conditions. Panel [C] addresses this issue.
The panel presents estimates in which we include controls for both the level and change
in the unemployment rate in the state where each newspaper is based. For completeness,
and because of potential crowding out effects in news coverage, we also control for the local
unemployment rate in the regressions for the three other economic issues. Again, the results
are similar to those reported in Table 3.
In panel [D] and panel [E] we control for various demographic variables in each newspa-
per’s market area that are correlated with newspaper readership and may also be correlated
with partisanship. In the table we focus on education (percent who graduated from college),
per-capita income, and percent living in urban areas. We do this by including the three-way
18
interaction terms involving these demographics, i.e. the interaction of the demographic con-
trol with EV it ·DPt. Once again, the pattern of estimates is similar to that in the baseline
specification. This further increases our confidence that the results in Table 3 are not driven
by omitted variable bias.29
In constructing NEj for Table 3 we include races in which a newspaper explicitly refused
to endorse any candidate, setting Ekjt = 0. This only happens 2.18 percent of the cases, but
these could be especially salient. Panel [F] shows what happens if we treat these as missing,
and simply drop them when estimating NEj. The results are again similar to those in Table
3.
Finally, panel [G] presents estimates where the dependent variable are defined in terms
of the total numbers of articles rather than the relative frequencies. We cannot directly
compare the coefficients with those in Table 3 since the scales are now different, but we see
that for unemployment the estimated coefficient on the three-way interaction term involving
endorsement partisanship is large and statistically significant at the .01 level. The results
for the budget deficit are slightly weaker, but even there the estimated coefficient on the
three-way interaction term involving reader partisanship is large and statistically significant
at the .10 level.30
To reiterate, we conclude from Table 5 that the estimates in Table 3 are not fragile, but
generally stable across a variety of alternative specifications.
7. Discussion and conclusions
In this paper we have analyzed the relationship between reader and endorsement parti-
sanship of U.S. newspapers and the coverage of economic issues, as a function of the true
economic datum and the political affiliation of the incumbent president. We began with
29We thank an anonymous referee for this suggestion.30We ran two more robustness checks for the unemployment issue, not reported in Table 5. First, it is not
clear a priori whether levels or changes in economic variables are more newsworthy in the eyes of editors,journalists and readers. Thus, we also estimated equation (3) using the change in unemployment ratherthan the level in the interaction term. The estimates using changes are small and statistically insignificant.Second, the dependent variables in Table 3 include both new articles and editorials. We constructed anotherdependent variable that excludes all stories containing the keyword “editorial” or “editor.” The estimatedcoefficient on the three-way interaction term is about 11% smaller than those in Table 3, and statisticallysignificant at the .05 level. Thus, a large part of the differential coverage of unemployment takes place onthe news pages, not merely on the editorial pages, suggesting that agenda-setting indeed spills over into theeconomic news section.
19
a case study of the Los Angeles Times, and find a striking correlation between the biases
in coverage of news on unemployment and inflation, and the partisan bias in political en-
dorsements on the editorial page. Next, we studied a large number of newspapers over the
period 1996-2005. There is strong evidence that newspapers endorsing Democratic candi-
dates give less coverage to high unemployment (and more coverage to low unemployment)
under Clinton than under George W. Bush, as compared to Republican-leaning newspapers.
This relationship is robust to a number of alternative specifications and robustness checks.
There is also evidence of a similar pattern of coverage for the trade deficit issue, although
this is less robust. For the budget deficit issue, we find evidence of a bias in coverage that
is correlated with the partisanship of newspapers’ readers, although this finding is not fully
robust. We find no significant patterns for inflation. Together, these finding suggest that
supply-side and demand-side forces both play a role in determining the bias in newspaper
coverage.
As mentioned in the introduction, we only study agenda-setting and do not attempt
to estimate any framing of economic events done through tone. Another limitation of our
approach is that we simply count the number of articles featuring the chosen keywords.31
One of the most desirable features of our approach is that it is quite flexible and eas-
ily replicable. This will allow us – and others – to readily extend the analysis in several
directions. First, we would like to expand the time window under consideration, for two
reasons. First, this would allow us to study more than two presidencies. Second, if we could
go back to the 1970s we would add an era where inflation was high and perceived as a much
bigger problem in our sample. Historical electronic archives such as ProQuest can be used to
construct long time series on the coverage of economic issues at least for a few newspapers.
Secondly, any debate on the extent of “mass media bias” in the U.S. should be put into
a comparative perspective.32 Given that the economy represents a salient issue in almost
all countries, one could use the same keywords-based search procedure on the electronic
31One could for example refine the search algorithm to code the page number and newspaper section onwhich each piece appears. In particular, one could give a higher weight to front page stories, or separatelyconsider them in the analysis. A further improvement (which is more difficult to implement within anautomated search) would be to weight articles by their length.
32See Gentzkow et al. [2006] for a time-series comparison of the extent of bias on the U.S. press in thecoverage of two political scandals, the Credit Mobilier in the 1870s and the Teapot Dome in the 1920s.
20
archives of newspapers and media outlets in other countries, and construct similar datasets
to the one analyzed here. The purpose of such an exercise would be to compare – on a cross-
country basis – the amount of within-country variation in the differential coverage of relevant
economic figures, as a function of the political affiliation of the incumbent government and
the level itself of the economic figure.
21
References
[1] Anderson, S. P. and McLaren, J. [2009]. “Media Mergers and Media Bias with Rational
Consumers”. Mimeo, University of Virginia.
[2] Ansolabehere, S., Lessem, R. and Snyder, J. M., Jr. [2006]. “The Orientation of Newspa-
per Endorsements in U.S. Elections, 1940-2002.” Quarterly Journal of Political Science,
1(4): 393-404.
[3] Antweiler, W. and Frank, M. Z. [2005]. “Do US Stock Markets Typically Overreact
to Corporate News Stories?” Mimeo, Sauder School of Business, University of British
Columbia.
[4] Baron, D. P. [2006]. “Persistent Media Bias”. Journal of Public Economics, 90(1): 1-36.
[5] Bernhardt, D., Krasa, S. and Polborn, M. K. [2008], “Political Polarization and the
Electoral Effects of Media Bias”. Journal of Public Economics 92: 1092-1104.
[6] Besley, T. and Prat, A. [2006]. “Handcuffs for the Grabbing Hand? Media Capture and
Government Accountability”. American Economic Review, 96(3): 720–736.
[7] Cameron, A. C., Gelbach, J. and Miller, D. [2008]. “Bootstrap-Based Improvements for
Inference with Clustered Errors”. Review of Economics and Statistics, 90: 414-427.
[8] Cohen, B. [1963]. The Press and Foreign Policy. Princeton, Princeton University Press.
[9] Corneo, G. G. [2006]. “Media Capture in a Democracy: The Role of Wealth Concen-
tration”. Journal of Public Economics 90: 37-58.
[10] DellaVigna, S. and Gentzkow, M. A. [2010]. “Persuasion: Empirical Evidence”. Annual
Review of Economics, 2: 643-670.
[11] DellaVigna, S. and Kaplan, E. [2007]. “The Fox News Effect: Media Bias and Voting”.
Quarterly Journal of Economics, 122: 1187-1234.
[12] Demsetz, H. and Lehn, K. [1985]. “The Structure of Corporate Ownership: Causes and
Consequences”. Journal of Political Economy, 93(6): 1155-1177.
22
[13] Durante R. and B. Knight [2010].“Partisan Control, Media Bias, and Viewer Responses:
Evidence from Berlusconi’s Italy”, Journal of the European Economic Association,
forthcoming.
[14] Eisensee, T. and Stromberg, D. [2007]. “News Floods, News Droughts, and U.S. Disaster
Relief.” Quarterly Journal of Economics, 122(2): 693-728.
[15] Ellman, M. and Germano, F. [2009]. “What do the Papers Sell? A Model of Advertising
and Media Bias”. Economic Journal, 119: 680-704.
[16] Erikson, R. S. [1989]. “Economic Conditions and The Presidential Vote”. American
Political Science Review, 83(2): 567-573.
[17] Erikson, R. S. [1990]. “Economic Conditions and the Congressional Vote: A Review of
the Macrolevel Evidence”. American Journal of Political Science, 34(2): 373-399.
[18] Fair, R. C. [1978]. “The Effect of Economic Events on Votes for President”. Review of
Economics and Statistics, 60(2): 159-173.
[19] Gasper, J. T. [2007]. “Ideological Shift: Explaining the Liberal
Media Myth.” Mimeo, Carnegie Mellon University. Available at
http://www.andrew.cmu.edu/user/gasper/WorkingPapers/Rep.pdf.
[20] Gentzkow, M. A., Glaeser, E. L. and Goldin, C. [2006]. “The Rise of the Fourth Estate:
How Newspapers Became Informative and Why it Mattered”. In Edward L. Glaeser
and Claudia Goldin (Eds.), Corruption and Reform: Lessons from America’s History.
National Bureau of Economic Research.
[21] Gentzkow, M. A. and Shapiro, J. M. [2006]. “Media Bias and Reputation.” Journal of
Political Economy, 114(2): 280-316.
[22] Gentzkow, M. A. and Shapiro, J. M. [2010]. “What Drives News Media Slant? Evidence
from U.S. Daily Newspapers.” Econometrica, 78(1): 35-71.
[23] Gentzkow, M. A., Shapiro, J. M. and Sinkinson, M. [2009]. “The Effect of Newspaper
Entry and Exit on Electoral Politics.” Mimeo, Chicago Booth School of Business.
23
[24] Gerber, A., Karlan, D. and Bergan, D. [2009]. “Does the Media Matter? A Field Exper-
iment Measuring the Effect of Newspapers on Voting Behavior and Political Opinions.”
American Economic Journal: Applied Economics 1(2): 35-52.
[25] Gilens M. and C. Hertzman [2000] “Corporate Ownership and News Bias: Newspaper
Coverage of the 1996 Telecommunications Act.” Journal of Politics, 62:369-386.
[26] Groseclose, T. and Milyo, J. [2005]. “A Measure of Media Bias”. Quarterly Journal of
Economics, 120(4): 1191-1237.
[27] Halberstam, D. [2000]. The Powers that Be. Chicago, University of Illinois Press.
[28] Hibbs, D. A, Jr. [1987]. The American Political Economy: Macroeconomic and Electoral
Politics in the United States. Cambridge, Harvard University Press.
[29] Iyengar, S., Kinder, D. R., and Peters, M. D. [1982]. “Experimental Demonstrations
of the ‘Not-So-Minimal’ Consequences of Television News Programs”. The American
Political Science Review, 76(4): 848-858.
[30] Iyengar, S. and Simon, A. F. [2000]. “New Perspectives and Evidence on Political Com-
munication and Campaign Effects”. Annual Review of Psychology 51: 149-169.
[31] Larcinese, V. [2007]: The Instrumental Voter Goes to the News-Agent: Information
Acquisition, Marginality, and the Media, Journal of Theoretical Politics, 19(3): 249-
276.
[32] Larcinese, V., Puglisi, R. and Snyder, J. M., Jr. [2007]. “Partisan Bias in Economic
News: Evidence on the Agenda-Setting Behavior of U.S. Newspapers.” National Bureau
of Economic Research working paper no. 13378.
[33] Lott, J. R., Jr. and Hassett, K. A. [2004]. “Is Newspaper Coverage of Economic Events
Politically Biased?” Working Paper, American Enterprise Institute, Washington, DC.
[34] MacKuen, M. B., Erikson, R. S. and Stimson, J. A. [1992]. “Peasants or Bankers? The
American Electorate and the U.S. Economy”. American Political Science Review, 86(3):
597-611.
24
[35] McCombs, M. E. and Shaw, D. L. [1972]. “The Agenda-Setting Function of Mass Me-
dia”. Public Opinion Quarterly, 36(2): 176-187.
[36] Mullainathan, S. and Shleifer, A. [2005]. “The Market for News”. American Economic
Review, 95(4): 1005-1030.
[37] Murphy, K. M. and Topel, R. H. [2002]. “Estimation and Inference in Two-Step Econo-
metric Models”. Journal of Business & Economic Statistics, 20(1): 88-97.
[38] Pagan, A. [1984]. “Econometric Issues in the Analysis of Regressions with Generated
Regressors”. International Economic Review, 25(1): 221-247.
[39] Petrocik, J. R. [1996]. “Issue Ownership in Presidential Elections, with a 1980 Case
Study”. American Journal of Political Science, 40(3): 825-850.
[40] Petrova, M. [2008a] “Inequality and Media Capture”. Journal of Public Economics, 92
(1-2): 183-212.
[41] Petrova, M. [2008b] “Mass Media and Special Interest Groups”. Mimeo, New Economic
School.
[42] Prat, A. and Stromberg, D. [2010]. “The Political Economy of Mass Media”. Mimeo,
London School of Economics.
[43] Puglisi, R. [2004]. “The Spin Doctor Meets the Rational Voter: Electoral Competi-
tion with Agenda-Setting Effects”. Available at SSRN: http://ssrn.com/abstract =
581881.
[44] Puglisi, R. [2006]. “Being the New York Times: the Political Behaviour of a Newspaper”.
Political Economy and Public Policy (PEPP) Working Paper n. 20, STICERD, London
School of Economics.
[45] Puglisi, R. and Snyder, J. M., Jr. [2008]. “Media Coverage of Political Scandals.” Na-
tional Bureau of Economic Research working paper no. 14958.
25
[46] Puglisi, R. and Snyder, J. M., Jr. [2009]. “The Centrist U.S. Press.” Mimeo, Department
of Political Science, MIT.
[47] Stromberg, D. [2004]. “Mass Media Competition, Political Competition, and Public
Policy”. Review of Economic Studies, 71(1): 265-284.
[48] Tufte, E. R. [1978]. Political Control of the Economy. Princeton, Princeton University
Press.
26
Figure 1: Dynamics of Democratic vote in California and LA Times endorsements
year2
Net Dem Endorsements Avg Dem Vote Pct - .5
1940 1950 1960 1970 1980 1990 2000
-1
-.5
0
.5
Figure 2: Coverage of unemployment and inflation on the LA Times.
Los Angeles Times, Pre- and Post- Otis Chandler
(Unem
plo
ym
ent H
its)/
(Tota
l H
its)
Early Period, 1948-1965Unemployment Rate
Actual hits (Repub=0, Democ=1) Predicted hits, Republican Pres Predicted hits, Democratic Pres
.03 .04 .05 .06 .07 .08
0
.005
.01
1
1
1
1
1 1
1 11
11
1
1
1
1
1
1
1
11
1
1
1 1
1
1
1
1
11
1
11
11 1
1
1
1
1
11
11
1
11
1
1
1
1
1
1
1
1
1
11
1
1
0
0
0 00
0 00
0
0
00
0
0
0
0
00
0
0
0
0
00
0
00
0
00
0 0
0
0
0
00
0
00
0
0
0
0
00
00
0
00
00 000
00
0
0 0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0 00
00
0
0
000
0 0
0
00
0
0
1
1
1
1
1
11
1
1
1
1
1
11
11
11
1
1
1
1
1
1
1
11
11
1
1
1
11
1
11
1
1
11
1
1
1111 1
1
1
1
1
1
1
1
1
1
1
1
1
(Unem
plo
ym
ent H
its)/
(Tota
l H
its)
Later Period, 1966-2005Unemployment Rate
Actual hits (Repub=0, Democ=1) Predicted hits, Republican Pres Predicted hits, Democratic Pres
.03 .04 .05 .06 .07 .08
0
.005
.01
.015
1
1
1
1
1
1
1
1
11
1 1
1
1111
1
111
1111
1
1
11
11
1
1
111
00000
0
00
0
0
0
00 0
00
0
0
00
0
0
0
0
00
0
0
0
00
0
0
0
00
00
0
0
0
0
0
0
0 00
00 0
00
00
0
0
00 0
0
0
0
0
0
00
0
0
0
0 0
0
0
0
0
00
00
0
0
0
0
0
1 1
1
11 1
1
1
1
1
1
1
1
1
11
1
1
1
1 1
1
1
111
1
111
11
11
1
1
1
1
1 1
1
1 11
1
1
1
1
0
0
00
00
0
0
0
0
0
0
0
0
0
000
0 00
00
0
0 0
0
0
00
00
00
0
0
0
00
000 0
00
00
0
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
00
0
0
0
00
0
0
00
0
00
0
00 00
0
00
0
0
0
0
0
0
0
0 0
0 0
0
00
0
000
0
0
0
00
1
1
1
1
1
1
1
1
1 1
1
1
1
11
1
111
1
11
1
1
1
1
11
11
1
1
1
111
11
1
1 1
1
1
11
1
1
1
1
1
11
1
1
1
1
1
1
111
1
1
1
111
11
111
1
1
1
11
1
1
1
11
11
1
1
11
11
1
1
1
11
100 0
00
0
0
0
0 00
00
00
0
000
0000
0
0
0
00
0
0
0
00
00
0
0
0
00
00
0
0
0
0
0
00 00
0000
0
0
0 0
0
(Inflation H
its)/
(Tota
l H
its)
Early Period, 1948-1965Inflation Rate
Actual hits (Repub=0, Democ=1) Predicted hits, Republican Pres Predicted hits, Democratic Pres
0 .01 .02 .03 .04
0
.005
.01
111
1
1
11
1
11
1
1
11
1
1
1
1
1
1
1
1
1
1
11
0
0
0
0
00
0 00
00
0
0
0
0000
00
0
00
0
0
000
0
00
0
0
00
0
0
0
0 0
00 0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
000
00
0
0
0
0
0
0
00
00
0
0
0
0
0
0 0
00 0
0
0
0000
00
0
0
0
0
1
1
1
1
11
1
1 1
1
11
1
11
1
11
11
1
111
1
1
1
11
111
1
1
1
111
11
1
1
11
11111
1 1
1
1
1
11
11
1
1
(Inflation H
its)/
(Tota
l H
its)
Later Period, 1966-2005Inflation Rate
Actual hits (Repub=0, Democ=1) Predicted hits, Republican Pres Predicted hits, Democratic Pres
.02 .04 .06 .08 .1 .12
0
.01
.02
.03
.04
1 11
1
1
1 1
1
11
1
11
1111111 1
1 11
1
1111 1
111111
00 0
0
0
0
0
000 0
0
00
0 0
00
00 00
00
00
00
0
0
0
0
00
000
000
00 00
0 0 0
0 0
00 0
0
00 0
0
0
0 0
0 0
0
0
0
00
0
0
00
00
0
0
00
0
000
0
00
00
00 0
0000
11
1
1
1
1
1
11
1
11
1
1
1
1
1
1 1 1
1
11
1
1
11
1
1
1
1
1
00
0
0
0
0
0
0
0
000
0
00
00
00
00
0
0
0
0
0
0
00
0
0000
0
0
0 0
00
0
00000
000
0
000
00
0
00
0
0
0
0
0
00 000
0
0
00
0
0 0
0
0
0
0
00
0
00
0
0
00
0
0
0
0
000 0
0
0 0
0
0
000
0
0
0
0
00
0
0
0
0 0
0
0
00
00
00
000
000
00000 0
000
00
0
000
111
111
1
11
11
111
11
1
11 1
1
1
11
1
11 1 1
111
1
1
11
1
111
11 1
111
1
1
1
1111
1
1
111
1
11
111 1
1
1
1
1
1
1
11
1 1 1
1
1
1
1
1
1
11
11
11
1 1
1
11111
00
0 0 00
0
0000
0
0
00
000
00
00
00
0 0
00
0000
00
0000
0
0
0
0
0
0
0
00
000
0
0
00
0
00
0
00
AASB
AC
ADSB
AGCB
AJ
AK
AS
AT
BD
BEBG
BI
BK
BL
BN
BNHB
BSCA
CDMB
CHTB
CIZB
CK
CL
CLDB
CO
CPDB
CR
CS
CSTB
DDNB
DM
DN
DPEC
ET
FB
FP
FTUB
FW
GB
GPTB
HC
HCBF
HDNB
HRNB
HT
IG
JG
JR JS
JW
KC
KX
KYPB
LA
LAT
LB
LH
LJSB
LVRB
MBRB
MH
MN
MS
MT
MWSB
NHRB
NJ
NT
NWDB
NYT
OK
OR
PBPB
PGPI
RM
RO
RTDB
SA
SAEC
SB
SE
SF
SFCB
SL
SP
SPTB
ST
TBTD
TLWB TNTB
TP
TT
UL
VC
WB
WE
WO
WP
WT
-1.5
-1-.
50
.51
en
do
rsem
ent p
art
isa
nship
.4 .5 .6 .7
reader partisanship
Fig. 3: Correlation between endorsement and reader partisanship
Table 1: variable definitions
symbol variable definition source
Unemployment U.S. monthly unemployment rate BLS, LNS 14000000
Inflation Monthly inflation rate, on annual basis BLS, CPI data, CUUR0000SA0
Budget deficit Quarterly federal deficit, as percentage of GDP BEA: NIPA Tables 3.2 and 1.1.5
Trade deficit Quarterly trade deficit, as percentage of GDP BEA: NIPA Tables 4.1 and 1.1.5
Relative frequency of unemployment stories Relative frequency of unemployment stories during month t on newspaper j electronic search on www.NewsLibrary.com: (unemployment OR jobless)
Relative frequency of inflation stories Relative frequency of inflation stories during month t on newspaper j electronic search on www.NewsLibrary.com: (inflation)
Relative frequency of budget deficit stories Relative frequency of budget deficit/surplus stories during quarter t on
newspaper j
electronic search on www.NewsLibrary.com: "government debt" OR
"government surplus" OR "government deficit" OR "federal debt" OR
"federal surplus" OR "federal deficit"
Relative frequency of trade deficit stories Relative frequency of trade deficit/surplus stories during quarter t on
newspaper j
electronic search on www.NewsLibrary.com: ("trade balance" OR "trade
deficit" OR "trade surplus")
U
jtn
B
jtn
T
jtn
I
jtn
UtEV
ItEV
BtEV
TtEV
Table 2: summary statistics, 1996-2005
symbol variable Obs. Mean Median Std. Dev. Min MaxMonthly unemployment rate 120 5.013 5.100 0.672 3.800 6.300
Monthly inflation rate 120 2.514 2.579 0.759 1.067 4.687
Quarterly budget deficit 40 1.047 1.229 1.936 -2.209 4.114
Quarterly trade deficit 40 3.432 3.604 1.579 1.070 6.166
Relative frequency of unemployment stories on newspaper j during month t 12004 0.689 0.633 0.372 0 3.138
Relative frequency of inflation stories on newspaper j during month t 12004 0.564 0.474 0.394 0 3.824
Relative frequency of budget deficit stories on newspaper j during quarter t 4009 0.123 0.102 0.099 0 1.887
Relative frequency of trade deficit stories on newspaper j during quarter t 4009 0.056 0.039 0.059 0 0.539
NEj Endorsement partisanship of newspaper j 101 0.029 0.128 0.340 -0.748 0.734
NRj Reader partisanship of newspaper j 101 0.517 0.524 0.074 0.384 0.689
Notes: all economic figures and relative frequencies of stories are expressed in percentage points.
j tn j tn j tn j tn U tx I tx B tx T tx
U
jtn
B
jtn
T
jtn
I
jtn
UtEV
ItEV
BtEV
TtEV
Table 3: Reader partisanship, endorsement partisanship and agenda bias in the coverage of unemployment and inflation
unemployment inflation budget deficit trade deficit(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Democratic President dummy x Economic variable x Reader partisanship - -0.264 -0.105 - -0.076 -0.088 - -0.065** -0.056** - -0.037 -0.017
[0.291] [0.300] [0.132] [0.158] [0.025] [0.028] [0.050] [0.046]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.122** - -0.115** 0.003 - 0.009 -0.01 - -0.006 -0.015* - -0.014*
[0.052] [0.053] [0.031] [0.036] [0.007] [0.008] [0.009] [0.008]
Ln of total articles 0.028 0.029 0.028 0.032** 0.032** 0.032** -0.02 -0.02 -0.02 0.004** 0.004* 0.004**
[0.022] [0.023] [0.022] [0.015] [0.015] [0.015] [0.022] [0.022] [0.022] [0.002] [0.002] [0.002]
Newspaper fixed effects yes yes yes yes yes yes yes yes yes yes yes yes
Date dummies yes yes yes yes yes yes yes yes yes yes yes yes
Newspaper-specific slope w.r.t. Economic variable yes yes yes yes yes yes yes yes yes yes yes yes
Newspaper-specific slope w.r.t. Democratic President dummy yes yes yes yes yes yes yes yes yes yes yes yes
Observations 12004 12004 12004 12004 12004 12004 4009 4009 4009 4009 4009 4009
Number of newspapers 101 101 101 101 101 101 101 101 101 101 101 101
R-squared 0.63 0.63 0.64 0.72 0.72 0.72 0.61 0.62 0.62 0.72 0.72 0.72
Notes: the table displays the output of fixed-effects regressions, with the relative frequency of stories about unemployment, inflation the budget and the trade deficit as the dependent variable. Observations are at the monthly level for unemployment and
inflation, and at the quarterly level of the two deficits.
Reader partisanship is the circulation-weighted Democratic partisanship of voters for each newspaper. Endorsement partisanship is the newspaper-specific propensity to endorse Democratic vs. Republican candidates. The focus is on the triple
interaction between the reader partisanship variable, the Democratic president dummy and the relevant economic variable, and on a similarly defined triple interaction with the endorsement partisanship variable. Standard errors are clustered at the
newspaper level and are reported in brackets below each coefficient. * significant at 10%; ** significant at 5%; *** significant at 1%
Table 4: Agenda bias in the coverage of economic news, implied magnitudes
[A] unemployment news Pro-Republican Endorsing Neutral Endorsing Pro-Democratic Endorsing
[B] budget deficit news Pro-Republican Readers Neutral Readers Pro-Democratic Readers
[C] trade deficit news Pro-Republican Endorsing Neutral Endorsing Pro-Democratic Endorsing
-0.023
9.78%
-19.29% -24.04% -28.44%
0.007 -0.002 -0.006
differential change in the relative frequency of news for a 1% increase in
unemployment under Democratic vs. Republican president
differential change in the relative frequency of news for a 1% increase in
budget deficit under Democratic vs. Republican president
differential change in the relative frequency of news for a 1% increase in
trade deficit under Democratic vs. Republican president
Notes: The first row in Panel [A] displays -for first, the third and the last quintile in the endorsement partisanship score- the differential change in the relative frequency of unemployment news for a one percent
increase in the unemployment rate under a Democratic vs. a Republican president. In the second row of the panel this differential change is reported as a percentage with respect to average relative frequency
of unemployment news in that quintile. Panel [B] replicates the format of Panel [A], but it focuses on budget deficit news and quintiles in the reader partisanship distribution. Finally, Panel [C] is devoted to
trade deficit news and quintiles in the endorsement score distribution.
-10.07%-2.87%19.56%
0.056 -0.014 -0.051
as percentage of average unemployment news for that quintile
as percentage of average budget deficit news for that quintile
as percentage of average trade deficit news for that quintile
-1.85% -7.15%
-0.035-0.029
Table 5: Agenda bias in the coverage of economic news, robustness checks
unemployment inflation budget deficit trade deficit
[A] bootstrapped standard errors
Democratic President dummy x Economic variable x Reader partisanship -0.105 -0.088 -0.056*** -0.017
[0.237] [0.127] [0.021] [0.035]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.115*** 0.009 -0.006 -0.014**
[0.041] [0.03] [0.006] [0.006]
[B] lagged value of economic controls
Democratic President dummy x Economic variable x Reader partisanship -0.139 -0.069 -0.026 -0.018
[0.284] [0.152] [0.029] [0.050]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.116** 0.013 -0.009 -0.014
[0.054] [0.036] [0.007] [0.009]
[C] controlling for state level unemployment
Democratic President dummy x Economic variable x Reader partisanship -0.122 -0.089 -0.054* -0.019
[0.306] [0.158] [0.028] [0.046]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.128** 0.009 -0.006 -0.014*
[0.053] [0.036] [0.008] [0.008]
[D] triple interactions with percent of college graduated and percent urban
Democratic President dummy x Economic variable x Reader partisanship -0.11 -0.092 -0.051* -0.004
[0.295] [0.179] [0.027] [0.044]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.109** 0.009 -0.006 -0.014
[0.055] [0.035] [0.008] [0.009]
[E] triple interactions with percent of college graduated, percent urban and income per capita
Democratic President dummy x Economic variable x Reader partisanship -0.117 -0.071 -0.049* 0.012
[0.288] [0.186] [0.027] [0.039]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.110* 0.011 -0.006 -0.012
[0.057] [0.035] [0.008] [0.008]
[F] dropping non-endorsements
Democratic President dummy x Economic variable x Reader partisanship -0.11 -0.089 -0.057** -0.017
[0.301] [0.159] [0.029] [0.046]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.107** 0.009 -0.005 -0.014*
[0.052] [0.036] [0.008] [0.008]
[G] count of articles as dependent variable
Democratic President dummy x Economic variable x Reader partisanship -9.674 -5.501 -3.914* -9.124
[9.328] [4.367] [2.290] [7.524]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -4.549*** -0.66 -0.345 -1.593
[1.373] [0.861] [0.592] [1.238]
Newspaper fixed effects yes yes yes yes
Date dummies yes yes yes yes
Newspaper-specific slope w.r.t. Economic variable yes yes yes yes
Newspaper-specific slope w.r.t. Democratic President dummy yes yes yes yes
Ln of total articles yes yes yes yes
Number of newspapers 101 101 101 101
Notes: the table displays the output of fixed-effects regressions, with the relative frequency of stories about the four economic issues as the dependent variable.
Reader partisanship is the circulation-weighted Democratic partisanship of voters for each newspaper. Endorsement partisanship is the newspaper-specific propensity to endorse
Democratic vs. Republican candidates. The focus is on the triple interaction between the reader partisanship variable, the Democratic president dummy and the relevant economic variable,
and on a similarly defined triple interaction with the endorsement partisanship variable.
In Panel [E] we use bootstrapped standard errors to account for generated regressors. In panel [B] we control for the lagged value of the economic variable, properly interacted with the other
variables of interest. In Panel [C] we control for the level and change of the unemployment rate in each state. In Panels [D] the Democratic president dummy and the economic variable are
interacted with the newspaper-specific, circulation-weighted percentage of college-graduated, percentage of individuals living in an urban area. In Panel [E] we also add the triple interaction
with income per capita. In Panel [F] we drop non-endorsements when calculating endorsement partisanship scores. In Panel [G] the dependent variable is the count of articles about a given
issue, instead of the relative frequency. Standard errors are clustered at the newspaper level and are reported in brackets below each coefficient. * significant at 10%; ** significant at 5%; ***
significant at 1%
Table A1: list of sampled newspapers
ID Newspaper State Chain Endorsement score Reader partisanship
AK Akron Beacon Journal OH Knight Ridder -0.1249327 0.5508947
AJ Albuquerque Tribune NM -0.1988737 0.5267518
AS Anchorage Daily News AK McClatchy Company 0.5296317 0.4139164
AT Atlanta Journal And Constitution GA Cox Newspapers 0.0776440 0.4723780
AGCB Augusta Chronicle GA Morris Communications -0.5251812 0.4305620
AASB Austin American Statesman TX Cox Newspapers 0.0690175 0.5289066
BS Baltimore Sun MD Tribune Co 0.1701819 0.5533579
BD Bangor Daily News ME 0.1972437 0.5341128
BE Bergen County Record NJ North Jersey 0.3031107 0.5004047
BI Birmingham Post Herald AL Advance Publications -0.4327879 0.4255932
BK Bismarck Tribune ND Lee Enterprises -0.1648864 0.3951091
BL Bloomington Pantagraph IL Lee Enterprises -0.5322713 0.4320979
BG Boston Globe MA New York Times 0.3423834 0.6142527
BNHB Boston Herald MA -0.3948854 0.6241412
BN Buffalo News NY 0.1331946 0.5813828
CR Cedar Rapids Gazette IA -0.2519723 0.5731747
CDMB Charleston Daily Mail WV Media News Group -0.5557753 0.5651314
CIZB Charleston Gazette WV 0.5179579 0.5651314
CO Charlotte Observer NC Knight Ridder 0.3377160 0.4406388
CSTB Chicago Sun Times IL Sun Times Media Group 0.0401971 0.6082023
CHTB Chicago Tribune IL Tribune Co -0.2390521 0.5605778
CK Cincinnati Post OH E.W. Scripps -0.3583446 0.3967437
CPDB Cleveland Plain Dealer OH Advance Publications -0.1711391 0.6064205
CS Columbia State SC Knight Ridder 0.1280723 0.4455652
CLDB Columbus Dispatch OH -0.4641224 0.4479681
CL Columbus Ledger-Enquirer GA Knight Ridder 0.3578027 0.5156434
OK Daily Oklahoman OK -0.4449803 0.4083460
DM Dallas Morning News TX Belo Corp -0.3165978 0.4298885
DDNB Dayton Daily News OH Cox Newspapers -0.1739211 0.4597971
NJ Daytona Beach News-Journal FL 0.6634993 0.5002059
DP Denver Post CO Media News Group 0.1464472 0.5093282
RM Denver Rocky Mountain News CO E.W. Scripps -0.2006290 0.5210701
FP Detroit Free Press MI Knight Ridder 0.1723710 0.5633393
NT Duluth News-Tribune MN Knight Ridder 0.1857737 0.6807154
ET Erie Times-News PA -0.3780103 0.5663361
EC Evansville Courier And Press IN 0.1223842 0.5055013
JG Fort Wayne Journal Gazette IN 0.1696575 0.3844016
FW Fort Wayne News-Sentinel IN Knight Ridder -0.3946047 0.3844016
ST Fort Worth Star-Telegram TX Knight Ridder 0.0619705 0.4318100
FB Fresno Bee CA McClatchy Company 0.2901540 0.4839964
GPTB Gary Post-Tribune IN Sun Times Media Group 0.1801603 0.5786577
GB Greensboro News And Record NC 0.3630669 0.4678081
HRNB Harrisburg Patriot-News PA Advance Publications -0.1866363 0.4087610
HC Hartford Courant CT Tribune Co 0.2975526 0.5803155
HDNB Hays Daily News KS 0.2304021 0.4167986
HCBF Houston Chronicle TX Hearst Corp -0.0600048 0.4503905
FTUB Jacksonville Florida Times-Union FL Morris Communications -0.7484340 0.4067213
KC Kansas City Star MO Knight Ridder 0.2187287 0.5233830
KYPB Kentucky Post KY E.W. Scripps 0.3707358 0.3967437
KX Knoxville News-Sentinel TN E.W. Scripps -0.3696161 0.4294882
LVRB Las Vegas Review-Journal NV Stephens Media Group -0.4533169 0.5084006
JW Lawrence Journal-World KS -0.4901071 0.5346774
LH Lexington Herald Leader KY Knight Ridder 0.7338427 0.4732455
LJSB Lincoln Journal Star NE Lee Enterprises -0.1099762 0.4753280Notes: the table contains the detailed list of sampled newspapers. For each of them we report the state where it is based, the chain to which it belongs (if any),
the endorsement partisanship score (NE_j) and the reader partisanship score (NR_j).
Table A1 (cont.): list of sampled newspapers
ID Newspaper State Chain Endorsement score Reader partisanship
LB Long Beach Press-Telegram CA Media News Group -0.3162732 0.5962189
NWDB Long Island Newsday NY Tribune Co 0.2694620 0.5621420
LA Los Angeles Daily News CA Media News Group -0.4658304 0.5931412
LAT Los Angeles Times CA Tribune Co 0.2755439 0.5450618
MT Macon Telegraph GA Knight Ridder 0.1543973 0.5185878
UL Manchester Union Leader NH -0.6915116 0.4686826
CA Memphis Commercial Appeal TN E.W. Scripps 0.2136075 0.5236294
MH Miami Herald FL Knight Ridder 0.2360750 0.5374542
MWSB Milwaukee Journal Sentinel WI 0.0769294 0.5340222
MN Minneapolis Star Tribune MN 0.3768361 0.5754108
MBRB Mobile Register AL Advance Publications -0.5841874 0.4035815
MS Modesto Bee CA McClatchy Company 0.0361410 0.4974961
NHRB New Haven Register CT Journal Register Co 0.1712770 0.5467662
TP New Orleans Times-Picayune LA Advance Publications 0.0687885 0.5845627
NYT New York Times NY New York Times 0.5000789 0.6444988
PBPB Palm Beach Post FL Cox Newspapers 0.3550132 0.5306050
JS Peoria Journal Star IL Copley Press -0.3065913 0.5143662
DN Philadelphia Daily News PA Knight Ridder 0.5311444 0.6723796
PI Philadelphia Inquirer PA Knight Ridder 0.2241287 0.5734859
PG Pittsburgh Post Gazette PA Block Family 0.1917282 0.6003192
OR Portland Oregonian OR Advance Publications 0.0550569 0.5745836
AC Press Of Atlantic City NJ -0.3603354 0.4983203
RTDB Richmond Times-Dispatch VA Media General -0.7237098 0.4277993
RO Roanoke Times VA Landmark Communications 0.4959546 0.4644936
SB Sacramento Bee CA McClatchy Company 0.6212453 0.5096645
SAEC San Antonio Express News TX Hearst Corp -0.1244709 0.4865679
SFCB San Francisco Examiner CA 0.2568252 0.6670643
SF Santa Fe New Mexican NM 0.1601401 0.6889591
SA Santa Rosa Press Democrat CA New York Times 0.2834142 0.6318817
HT Sarasota Herald-Tribune FL New York Times 0.1145749 0.4237741
IG Seattle Post-Intelligencer WA Hearst Corp 0.3295706 0.5903065
SE Seattle Times WA 0.0906542 0.5943558
JR Springfield State Journal-Register IL Copley Press -0.3228091 0.4862604
SL St. Louis Post Dispatch MO Pulitzer Inc 0.3797524 0.5507677
SP St. Paul Pioneer Press MN Knight Ridder 0.0880186 0.5994868
SPTB St. Petersburg Times FL 0.3527481 0.4913587
TNTB Tacoma News Tribune WA McClatchy Company 0.1705896 0.5550009
TD Tallahassee Democrat FL Knight Ridder 0.2836260 0.5478302
TT Tampa Tribune FL Media General -0.0701085 0.4641346
TB Toledo Blade OH Block Family 0.2734107 0.5557841
ADSB Tucson Arizona Daily Star AZ Pulitzer Inc 0.4503054 0.5405933
TLWB Tulsa World OK 0.1752578 0.4117864
VC Vancouver Columbian WA 0.1553285 0.5291076
WP Washington Post DC 0.2111574 0.6034141
WT Washington Times DC -1.1425110 0.6502780
WE Wichita Eagle KS Knight Ridder -0.2646916 0.4169850
WB Wilkes-Barre Times Leader PA Knight Ridder 0.2714555 0.5295009
WO Worcester Telegram And Gazette MA New York Times -0.4020534 0.5761151Notes: the table contains the detailed list of sampled newspapers. For each of them we report the state where it is based, the chain to which it belongs (if any),
the endorsement partisanship score (NE_j) and the reader partisanship score (NR_j).
Table A2: Agenda bias in the coverage of economic news, lagged economic variables and state-level unemployment rate
(1) (2) (3) (4) (5) (6) (7) (8)
Democratic President dummy x Economic variable x Reader partisanship -0.139 -0.122 -0.069 -0.089 -0.026 -0.054* -0.018 -0.019
[0.284] [0.306] [0.152] [0.158] [0.029] [0.028] [0.050] [0.046]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.116** -0.128** 0.013 0.009 -0.009 -0.006 -0.014 -0.014*
[0.054] [0.053] [0.036] [0.036] [0.007] [0.008] [0.009] [0.008]
State-level unemployment rate - 0.033 - -0.006 - -0.004 - -0.002
[0.021] [0.017] [0.004] [0.003]
Change in state-level unemployment rate - 0.071*** - 0.005 - 0.003 - 0.005*
[0.017] [0.013] [0.004] [0.003]
Ln of total articles 0.029 0.029 0.032** 0.032** -0.021 -0.02 0.004* 0.004*
[0.022] [0.022] [0.015] [0.015] [0.023] [0.022] [0.002] [0.002]
Newspaper fixed effects yes yes yes yes yes yes yes yes
Date dummies yes yes yes yes yes yes yes yes
Newspaper-specific slope w.r.t. Economic variable yes yes yes yes yes yes yes yes
Newspaper-specific slope w.r.t. Democratic President dummy yes yes yes yes yes yes yes yes
Observations 11996 12004 11996 12004 4001 4009 4001 4009
Number of newspapers 101 101 101 101 101 101 101 101
R-squared 0.63 0.64 0.72 0.72 0.61 0.62 0.72 0.72
unemployment inflation budget deficit trade deficit
Notes: the table displays the output of fixed-effects regressions, with the relative frequency of stories about the four economic issues as the dependent variable.
Reader partisanship is the circulation-weighted Democratic partisanship of voters for each newspaper. Endorsement partisanship is the newspaper-specific propensity to endorse Democratic vs. Republican
candidates. The focus is on the triple interaction between the reader partisanship variable, the Democratic president dummy and the relevant economic variable, and on a similarly defined triple interaction with
the endorsement partisanship variable. As a robustness check, in columns (1), (3), (5) and (7) we control for the lagged value of the economic variable (and its interactions) instead of the contemporaneous
one. In colums (2), (4), (6) and (8) we control for the level and change of the state-specific unemployment rate. Standard errors are clustered at the newspaper level and are reported in brackets below each
coefficient. * significant at 10%; ** significant at 5%; *** significant at 1%
Table A3: Agenda bias in the coverage of economic news, demographic controls
(1) (2) (3) (4) (5) (6) (7) (8)
Democratic President dummy x Economic variable x Reader partisanship -0.11 -0.117 -0.092 -0.071 -0.051* -0.049* -0.004 0.012
[0.295] [0.288] [0.179] [0.186] [0.027] [0.027] [0.044] [0.039]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.109** -0.110* 0.009 0.011 -0.006 -0.006 -0.014 -0.012
[0.055] [0.057] [0.035] [0.035] [0.008] [0.008] [0.009] [0.008]
Democratic Pres. dummy x Economic variable x Percent college-graduated -1.038*** -1.049*** 0.02 0.052 0.019 0.024 -0.086** -0.059
[0.347] [0.348] [0.193] [0.185] [0.026] [0.026] [0.040] [0.045]
Democratic Pres. dummy x Economic variable x Percent urban 0.398** 0.388** 0 0.028 -0.018 -0.014 0.005 0.027
[0.165] [0.187] [0.084] [0.091] [0.019] [0.020] [0.028] [0.031]
Democratic Pres. dummy x Economic variable x Income per capita - 0.001 - -0.003 - 0 - -0.002
[0.008] [0.003] [0.001] [0.002]
Ln of total articles 0.025 0.025 0.032** 0.032** -0.019 -0.019 0.005** 0.005**
[0.022] [0.022] [0.015] [0.015] [0.022] [0.022] [0.002] [0.002]
Newspaper fixed effects yes yes yes yes yes yes yes yes
Date dummies yes yes yes yes yes yes yes yes
Newspaper-specific slope w.r.t. Economic variable yes yes yes yes yes yes yes yes
Newspaper-specific slope w.r.t. Democratic President dummy yes yes yes yes yes yes yes yes
Observations 12004 12004 12004 12004 4009 4009 4009 4009
Number of newspapers 101 101 101 101 101 101 101 101
R-squared 0.64 0.64 0.72 0.72 0.62 0.62 0.73 0.73
Notes: the table displays the output of fixed-effects regressions, with the relative frequency of stories about the four economic issues as the dependent variable.
Reader partisanship is the circulation-weighted Democratic partisanship of voters for each newspaper. Endorsement partisanship is the newspaper-specific propensity to endorse Democratic vs. Republican
candidates. The focus is on the triple interaction between the reader partisanship variable, the Democratic president dummy and the relevant economic variable, and on a similarly defined triple interaction with
the endorsement partisanship variable. In columns (1), (3), (5) and (7) we control for the lagged value of the economic variable, properly interacted with the other variables of interest. In columns (2), (4), (6)
and (8) we control for the level and change of the unemployment rate in each state. Standard errors are clustered at the newspaper level and are reported in brackets below each coefficient. * significant at
10%; ** significant at 5%; *** significant at 1%
unemployment inflation budget deficit trade deficit
Table A4: Agenda bias in the coverage of economic news, dropping non-endorsements and count of articles as dependent variable
(1) (2) (3) (4) (5) (6) (7) (8)
Democratic President dummy x Economic variable x Reader partisanship -0.11 -9.674 -0.089 -5.501 -0.057** -3.914* -0.017 -9.124
[0.301] [9.328] [0.159] [4.367] [0.029] [2.290] [0.046] [7.524]
Democratic Pres. dummy x Economic variable x Endorsement partisanship -0.107** -4.549*** 0.009 -0.66 -0.005 -0.345 -0.014* -1.593
[0.052] [1.373] [0.036] [0.861] [0.008] [0.592] [0.008] [1.238]
total articles - 0.005*** - 0.004*** - 0.001*** - 0.000***
[0.001] [0.001] [0.000] [0.000]
Ln of total articles 0.028 0.302 0.032** 0.694 -0.02 1.141*** 0.004** -0.073
[0.022] [0.800] [0.015] [0.770] [0.022] [0.421] [0.002] [0.494]
Newspaper fixed effects yes yes yes yes yes yes yes yes
Date dummies yes yes yes yes yes yes yes yes
Newspaper-specific slope w.r.t. Economic variable yes yes yes yes yes yes yes yes
Newspaper-specific slope w.r.t. Democratic President dummy yes yes yes yes yes yes yes yes
Observations 12004 12004 12004 12004 4009 4009 4009 4009
Number of newspapers 101 101 101 101 101 101 101 101
R-squared 0.63 0.86 0.72 0.89 0.62 0.8 0.72 0.82
unemployment inflation budget deficit trade deficit
Notes: the table displays the output of fixed-effects regressions, with the relative frequency of stories about the four economic issues as the dependent variable.
Reader partisanship is the circulation-weighted Democratic partisanship of voters for each newspaper. Endorsement partisanship is the newspaper-specific propensity to endorse Democratic vs. Republican
candidates. The focus is on the triple interaction between the reader partisanship variable, the Democratic president dummy and the relevant economic variable, and on a similarly defined triple interaction with
the endorsement partisanship variable. In columns (1), (3), (5) and (7) we drop non-endorsements from the dataset. In columns (2), (4), (6) and (8) we use as dependent variable the count of articles about a
given economic issue, instead of the relative frequency. In this case we also control of the total number of articles per period. Standard errors are clustered at the newspaper level and are reported in brackets
below each coefficient. * significant at 10%; ** significant at 5%; *** significant at 1%
top related